基于知识图谱的目标战术意图识别仿真

Simulation of Target Tactical intention recognition based on knowledge Map

采用当前方法识别目标的战术意识图时,识别所用的时间较长,得到的识别结果与实际不符,存在识别效率低和识别准确率低的问题。将知识图谱应用于目标战术意识的识别中,提出基于知识图谱的目标战术意识图识别方法,构建战场目标的本体模型,分析战场目标本构模型中的二元关系,根据分析结果构建知识图谱。采集目标实时状态信息、目标属性信息和战场环境信息,采用经验模态分解方法对采集得到的数据信息做去噪处理,对去噪后的数据做归一化处理,将处理后的目标属性、战场环境和目标实时信息输入栈式自编码器中,得到一组输入信号,将输入信号输入知识图谱中实现目标战术意识图的识别。仿真结果表明,所提方法的识别效率高、识别准确率高。

When the current method was used to recognize the tactical consciousness graph of target, the recognition efficiency and the recognition accuracy were low. In this article, a method to recognize target tactical consciousness graph based on knowledge graph was proposed. First of all, the ontology model of battlefield target was constructed and the binary relationship of constitutive model of battlefield target was analyzed. According to the analysis result, knowledge graph was constructed. The real - time state information, target attribute information and battlefield environment information were collected. Moreover, the empirical mode decomposition method was used to reduce the noise of collected data. Then, the data after the noise reduction were normalized. In addition, the processed target attribute, battlefield environment and target real - time information were inputted into the stacked auto encoder, and then a set of input signals was obtained. Finally, the input signal was inputted into the knowledge map, so as to achieve the recognition of target tactical consciousness graph. Simulation results show that the proposed method has high recognition efficiency and high recognition accuracy.